3652 J . Org. Chem., Vol. 40, No. 25, 1975
Jung and Engel
Hydrogenolysis of Unsaturated Phosphate Esters Alfred Jung and Robert Engel* The City University of New York, Graduate Division, Department of Chemistry, Queens College, Flushing, New York 11367 Received May 27,1975
A series of unsaturated phosphate esters has been synthesized and subjected to hydrogenolysis over Adams catalyst. It has been demonstrated that direct hydrogenolysis of vinylic esters occurs yielding the phosphoric acid and the alkene; the alkene is then further reduced to the alkane. It has also been shown that migration of a "distant" olefinic linkage to a site subject to hydrogenolysis occurs at a rate such that cleavage product may be obtained. Moreover, it has been shown that benzylic ester linkages are subject to facile hydrogenolysis over Adams catalyst but not homobenzylic ester linkages.
It had previously been reported that diethyl isopropenyl phosphate (I) absorbed 2 mol of hydrogen over Adams catalyst yielding as t h e observed product diethylphosphoric acid; when palladium was used as t h e catalyst, only 1 mol of hydrogen was absorbed and diethyl isopropyl phosphate was observed t o be the pr0duct.l In the course of our investigations on t h e hydrogenolysis of aryl phosphate esters213 it was observed t h a t diethyl 1-cyclohexenyl phosphate (11) underwent hydrogenolysis over Adams catalyst with t h e formation of cyclohexane. With this observation it became of interest to investigate in detail the hydrogenolysis of saturated and unsaturated aliphatic phosphate esters. If a generality of reaction could be established here it would further extend t,he already well-known scheme of nonhydrolytic phosphate ester cleavages involving aryl and benzylic linkages.2-6 To test the generality of this reaction and t o elucidate the sequence of reaction steps the series of vinylic phosphate esters I-V was prepared and subjected t o atandard hydrogenolysis conditions. In all cases direct observation of hydrocarbon product and any intermediates was attempted. In addition, reactions using deuterium were perfprmed on the vinylic esters, their hydrocarbon cleavage products, and possible intermediates t o help establish the sequence of reaction steps. Several other molecular systems were also synthesized and subjected t o hydrogenolysis conditions. First, a fully saturated compound was investigated in a consideration of two possible reaction sequences, i.e., cleavage followed by reduction as compared t o initial reduction followed by cleavage. Moreover, as deuterium interchange was noted with possible intermediates containing sites of unsaturation, and as structural isomerization under hydrogenation conditions is well known,' it was of interest t o prepare and investigate a phosphate ester with a n olefinic site distant from a possible site of hydrogenolysis. Finally, benzylic and homobenzylic esters were investigated. Experimental Section Reagents. All reagents used in the preparation of the phosphate esters or their precursors were purchased from Aldrich Chemical Co. and used without further purification. Absolute ethanol, used as a standard solvent for the hydrogenolyses, was purchased from Commercial Solvents Corp. and used without further purification. The platinum oxide catalyst (83 f 0.5%)was a generous gift of Engelhard Minerals and Chemicals Corp. Deuterium was from Matheson Corp. and was in excess of 99.5 atom % D. General Synthesis of Diethyl Vinyl Phosphates. The standard reaction technique used for the Perkow reaction was fol10wed.*~~ To 0.20 mol of triethyl phosphite heated to 120" was added in small portions 0.20 mol of the appropriate a-chloro ketone. The reaction mixture was maintained at 120" for 1 hr after completion of the addition and then raised t o 170" for an additional 1 hr, whereupon the product was vacuum distilled. Yield and analytical data are given below for each compound. Satisfactory ir,
NMR, and mass spectra were obtained for all compounds, and elemental analyses for those not previously reported.1° Diethyl 2-Propenyl Phosphate (1):l yield 47%; bp 64" (0.9 Torr); mass spectrum, parent peak rnle 194 (Z6%), base peak rnle 99. * Diethyl 1-Cyclohexenyl Phosphate (11):9 yield 35%; bp 105O (0.15 Torr); mass spectrum, parent peak rnle 234 (54961,base peak mle 99. Diethyl 1-Cyclopentenyl Phosphate (111): yield 78%; bp 83' (0.09 Torr); mass spectrum, parent peak mle 220 (33%),base peak rnle 137. Anal. Calcd for CgH1704P: C, 48.88; H, 7.85. Found: C, 49.09; H, 7.73. Diethyl 1-Cycloheptenyl Phosphate (IV): yield 78%; bp 80" (0.01 Torr); mass spectrum, parent peak rnle 248 (40%),base peak rnle 155. Anal. Calcd for CllH2104P: C, 53.56; H, 8.53. Found: C, 53.23; H, 8.47. Diethyl a-Styryl Phosphate (V): yield 39%; bp 106' (0.04 Torr); mass spectrum, parent peak rnle 256 (37%),base peak mle 106. Anal. Calcd for C12H1704P: C, 56.06; H, 6.78. Found: C, 56.21; H, 6.64. General Synthesis of Diethyl Alkyl Phosphates. To 0.12 mol of the alcohol in 100 ml of dry benzene was added slowly 0.1 mol of sodium hydride (ether washed) and the solution was stirred for 2 hr. To this was added dropwise 0.1 mol of diethyl phosphorochloridate and the reaction mixture was stirred for several hours. There was then added 50 ml of pentane, causing precipitation of the salt, which was filtered with suction. The solvent was removed at reduced pressure and the residual phosphate was vacuum distilled. Yield and analytical data are given below for each compound. Diethyl Cyclohexyl Phosphato (VI):" yield 45963; bp 103O (0.65 Torr); mass spectrum, parent peak mle 236 (l%), base peak m/e 155. Diethyl Benzyl Phosphate (V11)P yield 25%; bp 101" (0.8 Torr); mass spectrum, parent peak mle 244 (67%),base peak rnle 91. Diethyl 2-Phenylethyl Phosphate (VI11)P yield 43%; bp 147' (0.25 Torr); mass spectrum, parent peak mle 256 (3%),base peak mle 106. Diethyl 6-Pent-1-enyl Phosphate (IX): yield 16%; bp 85" (0.75 Torr); mass spectrum, parent peak rnle 222 (12%),base peak m / e 99. Anal, Calcd for CgHlg04P: C, 48.65; H, 8.56. Found: C, 48.93; H, 8.64. General Procedure for 1 Atm and 4 Atm Hydrogen Reactions. The reactors and experimental conditions for hydrogenation were as previously de~cribed.~ For the analytical experiments the reaction flask was charged with 25 ml of absolute ethanol ca. 0.020 M in the compound to be investigated and ca. 0.020 M in a reference material for gas-liquid chromatographic (GLC) analysis. For the studies involving the isolation and identification of products the concentration of the reactant was increased to ca. 1M . Deuterium Incorporation Studies. All experiments using deuterium were performed at 1 atm pressure with solutions ca. 1 M in phosphate ester. The extent of deuterium incorporation was determined by collection of product or "unreacted" starting material using gas-liquid chromatography and its analysis by mass spectrometry. The resultant spectra were compared with those of the corresponding undeuterated materials. Analysis. Gas-liquid chromatographic analysis (and preparative GLC) were performed using two columns: a 10 ft X 0.25 in. colunn of 20% Carbowax 20M on Chromosorb W was used for analysis of reactions of 111 whereas a 5 f t X 0.26 in. column of 20% Apiezon L on Chromosorb W was used for all other reaction sys-
J. Org. Chem., Vol. 40, No. 25, 1975
Hydrogenolysis of Unsaturated Phosphate Esters tems. For quantitative analysis all products were compared for GLC relative response factors with a reference material; cyciohexane was used as the reference material for all reaction systems except those involving I1 and VI and the deuteration studies, where methylcyclohexanewas used as the reference material. All ir spectra were measured using a Perkin-Elmer Model 237-B spectrometer; NMR spectra were measured using a Varian EM360 spectrometer and mass spectra were measured using a Varian MAT CH-7 instrument.
Table I 0
II
(CH,CH,O),P-OR
H,, PtO,
(CH,CH,O),PO,H + R”
----+
Yield R’H(24 hr reaction time),a % R _
~
Compd -
-e
Results a n d Discussion The diethyl vinyl phosphates I-V were subjected to hydrogenation in ethanol solution over Adams catalyst a t 1 and 4 atm pressuie of hydrogen. The products and yields (as determined by GLC) are listed in Table I. With the exception of I, where a true quantitative determination could not be made owing to volitalization of the product from the reaction mixture, quantitative or near quantitative cleavage of the vinylic ester group was observed. This is interesting as it points to the extremely facile nature of the hydrogenolysis process as compared to other possible reactions, such as olefin reduction and hydrogen exchange (vide infra). With reaction systems of 11-V attempts were made to observe and isolate cleaved olefinic intermediate; GLC collection of the entire region of expected elution was performed and the effluent subjected to mass spectral analysis. No olefinic material could be observed even m d e r these extremely sensitive conditions of analysis. Changing to a saturated hydrocarbon solvent, useful in slowing the reduction of aro1natics,~,3 again proved unsuccessful. Either the free olefinic hydrocarbon is not an intermediate, or it is formed in only low concentrations and is reduced to the alkane with a high rate. Several experiments were performed to consider these possibilities. First, in a consideration of the possibility of initial reduction of the vinylic phosphate, the fully saturated system VI was prepared and subjected to the identical conditions as used for 11. After 1week using conditions of 1 atm pressure of hydrogen, diethyl cyclohexyl phosphate exhibited no cleavage and could be recovered unchanged. After 1 week under conditions of 4 atm pressure of hydrogen only 5% cleavage (to cyclohexane) could be observed. From this it may be concluded that initial reduction of the vinylic ester followed by hydrogenolysis does not occur. Second, in a consideration of the possibility of the free alkene being an intermediate, but present a t any one time in only low concentration, evidence of an indirect nature was gathered. Several experiments were performed using deuterium in deuterated alcohol (CH30D, CDsOD, CHZCH~OD)and cyclohexane solvents. In these solvents with Adams catalyst and deuterium it was observed that the alkenes to be considered as intermediates underwent significant deuterium incorporation in excess of that expected by addition to the olefinic linkage. T h a t this deuterium-for-hydrogen exchange occurred with the alkene and not its reduction product was shown by further experiments of the same type with the alkane itself; within experimental error no deuterium incorporation could be observed for the fully saturated compounds under these reaction conditions. Thereby, were the product alkane (from vinylic phosphate cleavage with deuterium) to exhibit incorporation of more than three atoms of deuterium it could be indicative of alkene intermediacy. For compound 11, studied in CH30D and CD30D solution, use of deuterium for the cleavage reaction resulted in formation of cyclohexane exhibiting various quantities of deuterium incorporation, up to seven atoms of deuterium in significant amount. Similarly, for compound IV, studied in cyclohexane solution, incorporation of u p to seven atoms of deuterium in
3653
-
_
I I1
l a t m H, _
~
I
_
>15h
A
@
97.6
_
4 atm H, --
>6b
fi
0
100.0
-0
I11
0
97.3
0
100.0
-0
IV
0
96.7
(c
100.0
/-Q
94.4
f
-c> &$
A
VI VI1
VI11
0 -c>
a
-
0
oc
@
8.6
-c>
60.0
a
0
5c 86.0
